To reduce the sidelobe level (SLL), a ten-element non-uniform microstrip array antenna at 14.25 GHz frequency was designed. This Letter proposes an improved differential evolution algorithm, expanding the optimisation variable from one, the array element spacing, to two, the array element spacing and the excitation amplitude. Then a series linear array antenna with non-uniform spacing and unequal excitation is constructed, and a uniform spacing Chebyshev array antenna and an equal excitation non-uniform spacing array antenna are constructed as references. The experimental results show an SLL of −22 dB and a gain of 16.5 dB.

A novel feedforward-coupling ring oscillator (FRVCO) is presented. A comparative study in terms of oscillation frequency and phase noise is conducted between the proposed model and the traditional FRVCO. Due to the improved delay cell, high oscillation frequency, low phase noise and uniform multi-phase output can be achieved. Compared with the traditional FRVCO, under the same design parameters, the FRVCO proposed in this Letter can increase the oscillation frequency by 10% while reducing the phase noise at 1 MHz by at least 10 dB.

This Letter presents a self-capacitance sensing (SCS) technique for touch screen panel with the alternating panel charge sharing (APCS) that effectively suppresses low-frequency noise and achieves a better signal-to-noise ratio (SNR). With the proposed APCS-based SCS, SNR is improved in measurement by 7.65 dB compared to the conventional panel charge sharing-based SCS.

Mobile robot path planning is an important part of the mobile robot field. The swarm intelligence optimisation algorithm has certain advantages in solving such multi-objective optimisation problems, so this Letter proposes to apply the pigeon-inspired optimisation (PIO) algorithm to the path planning problem. However, the traditional PIO algorithm is suffering from the problems of easy to get into local optimisation, low stability and premature convergence. To overcome its shortcomings, this Letter proposes a logistic beetle algorithm search–PIO (LBAS-PIO) algorithm. In the optimisation process of the LBAS-PIO algorithm, the PIO algorithm is initialised by the logistic mapping, making the search space more extensive. Learning from the idea of the beetle algorithm search, each pigeon has its own judgment of the environment and the number of iterations and search time are reduced. Further, the path evaluation function of the algorithm and the execution order of the operators are optimised to make the algorithm smoother and easier to find the global optimal solution. Simulation and experimental results illustrated the superiority of the LBAS-PIO algorithm and the LBAS-PIO algorithm better meets the needs of mobile robots for the optimal path planning.

Ferroelectric tunnel junctions have recently attracted a lot of research interest, not only for memory operation but also for e.g. neuromorphic computing. Retention characteristics of these devices are of utmost importance in many applications. The authors have recently demonstrated the feasibility of double-layered ferroelectric tunnel junctions where the ferroelectric hafnium zirconium oxide acts as the memory layer and the dielectric aluminium oxide as the tunnelling barrier. The depolarisation field from the unscreened polarisation charges that is introduced by this structure causes serious limitations on the retention behaviour due to the unintentional back-switching of ferroelectric domains. This leads to a diminishing memory window over time. By using a built-in bias field from different work function metal electrodes, in this case platinum and titanium nitride with 0.85 eV difference, one can shift the polarisation hysteresis curve and counteract the electric depolarisation field. This leads to a more stable on-current over time while only marginally increasing the off-current degradation. Extrapolation to ten years shows a much larger memory window compared to the symmetric metal electrode structure.

This Letter presents an alternative analysis method to derive the voltage–current transmission matrix of transmission line sections. By contrast with the approach usually adopted in many microwave engineering textbooks, the proposed method, instead of coping with the telegrapher's equations and their solutions, computes efficiently the relationship between the voltages and currents at the input and output of a circuit modelled through N identical cascaded two-port reciprocal networks. This is implemented through the eigendecomposition of the transmission matrix of the two-port networks which are subsequently synthesised with L-sections. Assuming N as being very large, one obtains the desired transmission matrix, and additionally, it is depicted that the circuit supports forward and backward waves.

RGB-D-based human action recognition aims to learn distinctive features from different modalities and has shown good progress in practice. However, it is difficult to improve the recognition performance through directly training multiple individual convolutional networks (ConvNets) and fusing features later because complmentary information between different modalities cannot be learned. To address this issue, this Letter proposes a single two-stream ConvNets framework for multimodality learning that extract features through RGB and depth streams. Specifically, the authors first represent RGB-D sequence to motion images as the inputs of the proposed ConvNets for obtaining spatial–temporal information. Then, a features fusion and joint training strategy is adapted to learn RGB-D complementary features simultaneously. Experimental results on benchmark NTU RGB+D 120 dataset validate the effectiveness of the proposed framework and demonstrate that two-stream ConvNets outperforms the current state-of-the-art approaches.

This Letter presents a method for training convolutional neural networks (CNNs) that detect targets of interest in hyperspectral images. Collecting suitable and abundant training data has been the main obstacle to the successful application of CNNs to hyperspectral target detection. To solve the problem, the authors propose a scheme to generate synthetic training data. Publicly available spectral reflectance library and an easy-to-obtain radiative transfer model are utilised in their scheme. Using the synthetic training data only, a Siamese CNN is trained to learn robust features for the detection task. Experimental results on the real hyperspectral image show the effectiveness of the proposed method.

In this study, the authors present a novel visual tracking method using the pixel-wise posterior estimation and minibatch Monte Carlo sampling. To avoid background pixels in the noisy bounding box representation, they estimate the posteriors in a pixel-wise manner. To boost the pixel-wise posterior estimation, they adopt minibatch Monte Carlo sampling, where only a small portion of pixels are used for inference. Experimental results demonstrate that the proposed visual tracker produces accurate tracking results using a small portion of pixels for the posterior estimation and is comparable to state-of-the-art methods.

This Letter presents an ultrasonic waveguide sensor for measuring fluctuations in a liquid level using leaky longitudinal wave L(0,2) in a coolant loop system. A 20% cold-worked D9 fuel pin clad tube of 1000 mm long, 6.6 mm outer diameter and 0.45 mm wall thickness was used as a waveguide on which the L(0,2) mode was transmitted/received along the axial direction at 500 kHz. When the waveguide got introduced into a coolant loop and partially immersed in a non-viscous liquid (water/liquid sodium), the amplitude of L(0,2) mode got reduced due to leakage. The rate of leakage depends on the characteristics of the waveguide and the embedding liquid medium. By utilising this physical property of L(0,2) mode, it is possible to predict the variations in liquid level during dynamic conditions. Experimental measurements were carried out on a mock-up test loop in which water was used as embedding liquid and two artificial notches were made on the waveguide at different depths were used as reference levels. The time-amplitude signals were monitored and the reflected energies from the bottom end of the waveguide and the notches were used as the threshold and reference amplitudes respectively for identification of water level.

A wide locking range (LR) injection locked frequency tripler (ILFT) based on a dual-band voltage controlled oscillator (VCO) is proposed in this Letter. By rationally designing the transformer parameters, the VCO can easily realise the adjacent dual-band operating characteristics. The ILFT has dual LRs which can easily overlap each other at a fixed bias. A subharmonic signal is also injected through the transformer for frequency locking. A wide LR of 27.1% from 69 to 90.6 GHz at an injection power of 0 dBm is achieved whereas the tripler only consumes 5.88 mW under a 0.6 V supply voltage.

This Letter proposes a high data rate optical transmission system for data centre interconnects, where the authors employ vertical cavity surface emitting lasers (VCSEL) and graded index few-mode fibre (GI-FMF) having modal bandwidth of around 24.3 GHz.km to enhance the link capacity and range. Additionally, they consider PAM-4 modulated data transmission at the rate of 2100 Gbps using a pair of VCSELs and employing mode group division multiplexing of two linearly polarised (LP) modes and of the GI-FMF. Digital signal processing techniques are applied at the receiver to process the received PAM-4 signals for equalisation in order to mitigate the channel impairments. Furthermore, standard OM3 and OM4 fibres are considered as a benchmark to compare the transmission performance of the designed GI-FMF. They observe that the designed GI-FMF supports longer transmission distance for both and modes upto 900 and 450 m at 50 and 100 Gbps data rates, respectively, while achieving lower receiver sensitivities for forward error correction limit of . The proposed optical transmission system can be a potential applicant in future data centre networks supporting transmission rates in 800 Gbps to 1.6 Tbps range.

This Letter presents the mechanical analysis of polymethyl methacrylate microstructured polymer optical fibres drawn at different pressures. The fibres were drawn at 200, 400 and 600 mmH₂O, where all the samples have a diameter close to 250 µm. Tensile tests were performed in five samples of each drawing pressure and the statistical analysis was applied to Young's modulus, yield stress, strain at breakage and Poisson's ratio of each sample. Results show a statistically significant increase of Young's modulus with the increase in drawing pressure, whereas the strain at breakage presented the opposite behaviour, with its values decreasing as the drawing pressure increases. Furthermore, the Poisson's ratio, as well as, the yield stress do not show a statistically significant variation with the different drawing pressure analysed.

In this Letter, the authors designed a photonic configurable logic block (pCLB) with Mach–Zehnder modulator as a universal photonic gate (UPG) using electro-optic mechanism. The UPG is capable of performing all the logical operations with the help of FPGA as a configuration controller. The functionality and its suitability is verified by forming a pCLB of ten UPGs, which finds applications in digital photonics. For the design of pCLB, traditional electronic design automation tools are used. The pCLB provides features namely scalability and modular approach, for the reconfigurable digital photonic integrated circuits. The photonic gate designed using a two-dimensional finite difference time domain approach shows a crosstalk-free operation with a negligible transmission loss of 0.18 dB and flexible switching control compared to the existing MZI-based photonic gate.

This Letter describes a pulsed waveform which exhibits extreme Doppler tolerance, enabling large time-bandwidth-product observations of high-velocity targets with low computational complexity. The waveform, which supports pulsewidth, pulse repetition interval, and frequency agility, utilises hyperbolic frequency-modulated pulses with modulation parameters constrained to suppress range migration.

The authors derive a closed-form exact expression and a fast convergent series for the probability of phased-array detection of non-fluctuating targets embedded in complex white Gaussian noise. To be realistic, they assume that the amplitude of the target echoes and the noise power are unknown parameters. Their series representation achieved impressive savings in computational load and computation time when compared to the numerical evaluation of the existing integral-form solution.

In this Letter, the authors report on a 26 GHz joint communication–radar system. By exploiting the beam steering features of leaky-wave antennas, angle-of-arrival estimation and multi-user transmission are achieved. Furthermore, the Zadoff–Chu preamble of the OFDM waveform is used to provide a frequency-modulated continuous-wave radar signal. This way, the wide operational bandwidth is used for both, communications and radar, yielding high data throughput and precise user localisation. Experimentally, spectral-efficient 16-QAM multi-user 5 × 4 Gbit/s communications and user localisation with an accuracy of about 2 cm and ±1.5° is demonstrated.

This Letter reports a high-performance film bulk acoustic resonator (FBAR) that can be applied in 5G wireless communication. The FBAR has a back-etched free-standing structure using an ultra-thin single-crystalline lithium niobate (LiNbO₃) as the piezoelectric film. The thin LiNbO₃ was obtained by the smart cut method and FBARs were fabricated by a standard MEMS process. Owing to the superior bulk-like properties of the single-crystal thin film, the fabricated FBARs have a resonant frequency of 5.0 GHz, a quality factor above 1800 and an electromechanical coupling coefficient of 2.66%, demonstrating a promising potential of FBAR filters in 5G and future 6G applications.

Automatic modulation classification (AMC) is an important and challenging task that aims to discriminate modulation formats of received signals, such as military communications, cognitive radio and spectrum management. With the development of deep learning techniques, research in AMC has gained promising results because of its powerful representation and classification abilities. In this Letter, the authors present a new network architecture that combines a frequency selection module and a convolutional neural network (CNN). This scheme not only processes raw signal data with carriers to increase the in-band signal-to-noise ratio but also converge faster than traditional CNN. Experiments demonstrate the effectiveness and efficiency of the proposed model.

To decrease the network size of quantised kernel least mean square (QKLMS) dramatically, the QKLMS algorithm with an online learning vector strategy, which is named LV-QKLMS, is proposed in this Letter. The centres of the dictionary in LV-QKLMS are updated dynamically by the online learning vectors. Unlike QKLMS only updating the coefficient of the nearest centre to current input, LV-QKLMS updates the centres of the dictionary using the redundant input data accordingly. Thanks to the iterative learning vector, LV-QKLMS has a constantly changing quantisation area on the arrival of a new sample, and thus generates a superimposed quantisation area containing more redundant input data than QKLMS with a fixed quantisation area. In addition, the learning vector updated by a linear combination of input data and itself in the superimposed quantisation area guarantees the filtering accuracy of LV-QKLMS. Simulation results on the prediction of Mackey–Glass chaotic time series are presented to validate the effectiveness of LV-QKLMS.

In this Letter, a deep reinforcement learning-based approach is proposed for the delay-constrained buffer-aided relay selection in a cooperative cognitive network. The proposed learning algorithm can efficiently solve the complicated relay selection problem, and achieves the optimal throughput when the buffer size and number of relays are large. In particular, the authors use the deep-Q-learning to design an agent to estimate a specific action for each state of the system, which is then utilised to provide an optimum trade-off between throughput and a given delay constraint. Simulation results are provided to demonstrate the advantages of the proposed scheme over conventional selection methods. More specifically, compared to the max-ratio selection criteria, where the relay with the highest signal-to-interference ratio is selected, the proposed scheme achieves a significant throughput gain with higher throughput-delay balance.

As one of the 5G waveform candidates, generalised frequency division multiplexing (GFDM) systems have been proposed to achieve lower out of band radiation and stronger robustness against synchronisation errors. However, theoretical BER performances of GFDM systems with in-phase/quadrature (IQ) imbalance, as well as compensation strategies, have not been widely investigated. In this Letter, the authors evaluate theoretical BER performances of GFDM systems with transmitter IQ imbalance (IQI) before and after a proposed compensation method. Basically, they give approximated BER expressions for different modulations through considering introduced IQI completely as useless interference, and then derive a more precise BER expression for QPSK modulation. Additionally, they put forward a compensation method to remove IQI interference in GFDM systems at the cost of enlarged noise. Furthermore, analytical BER expressions with IQI compensation are derived, and then the effectiveness of the proposed compensation strategy is proven by theoretical and simulation results.

A novel method based on Euclidean algorithm is proposed to solve the problem of blind recognition of binary Bose–Chaudhuri–Hocquenghem (BCH) codes in non-cooperative applications. By carrying out iterative Euclidean divisions on the demodulator output bit-stream, the proposed method can determine the codeword length and generator polynomial of unknown BCH code. The computational complexity is derived as O(n ³). Simulation results show the efficiency of the proposed method.

Orthogonal frequency division multiplexing with index modulation (OFDM-IM) is a promising multicarrier transmission technique that exploits both conventional modulation and the indices of active subcarriers for data transmission. In this Letter, the authors propose OFDM with power distribution index modulation (OFDM-PIM) where a unique power distribution technique is applied to index modulated subcarriers. Furthermore, we propose OFDM with in-phase/quadrature PIM (OFDM-I/Q-PIM) which performs the same power distribution process on the real and imaginary parts of the data symbols, independently. Average bit error performances of OFDM-PIM and OFDM-I/Q-PIM are obtained and the superiority of the proposed schemes over reference schemes is demonstrated via extensive computer simulations.